IntroductionTCM recently began testing its industrial-size carbon capture facility. To meet the regulated emission criteria, it is of im-portance to qualitatively and quantitatively follow the discharge of amines and their degradation producuts to the atmos-phere during operation. The real-time, online quantification of such volatile/semivolatile organic compounds from stack emission is feasible with the deployment of Proton Transfer Reaction-Time of Flight-Mass Spectrometry (PTR-ToF-MS) due to its fast response, high sensitivity in the sub ppb range, wide dynamic range, and its unique mass resolution allowing formula determination.a The capabilities and uncertainties of the proposed method were determined using in-situ calibrati-on data and by comparing with neighbouring FT-IR analyzer during Aker Clean Carbon's (ACC) plant commissioning phase using an aqueous monoethanolamine (MEA) solution. Thus, the measurements presented here represent results from an initial, non-optimized carbon capture testing phase in August 2012.

Claus Nielsen1, Liang Zhu1, Gunnar Schade1, Anne Kolstad-Morken2, Sissel Nepstad2, Øyvind Ullestad2

1 Department of Chemistry, University of Oslo, Norway 2 Technology Center Mongstad, TCM, Norway

Conclusions

Results

These initial results suggest that long-term emission monitoring from full-scale carbon capture facilities using PTR-ToF-MS is possible, repro-ducible, and quantitative within 10 to 30 percent. The amine (MEA) pre-cursor and its major volatile degradation products were measured. Masses corresponding to potentially harmful minor byproduct emissions, e.g. ni-trosamines and nitramines, were not detected above their detection limits (low ppb levels). Further testing will be necessary to determine whether measurement biases or artefacts exist due to the current instrument, sam-pling line, or gas composition parameters.

The inflow to the stack, outflow from the absorber, and the one from stripper column were periodically directed, through the heated (90 °C) and constantly flushed Teflon PFA lines, to a hot stainless steel manifold, to which both PTR-ToF-MS and FT-IR analyzer were connected (Fig. 1). The main running parameters of PTR-ToF-MS (Ionicon) were set by drift tube temperature (110 °C), drift voltage (500 V) and drift tube pressure (2.2 mbar), corresponding to an E/N value of ~125 Td.

Methods

Figure 3: Monoethanolamine (MEA) measurements performed using PTR-ToF-MS, Finetech FT-IR, and YAGA Gasmet FT-IR on the top of the absorber tower. Sections “a, b, c” denote a peri-odic sampling sequence with a 45 min cycle: To Absorber, From Absorber, and Stripper, respectively. The non-quantitative CO2 trace was plotted to guide the eye on the sampling cycles. Note that there was no flue gas entering the absorber before 17:30 h when the carbon capture processes were reinitialized. Thus, un-realistically high amine mixing ratios occurred prior to that time, which enabled instrument comparisons. In the first 4 hours, the MEA signal in the PTR-ToF-MS corresponded reaso-nable well to the Finetech FT-IR and at times the YAGA analy-ser. The much larger discrepancies at later stages could be due to the increased uncertainties of FT-IR measurements in the sub-ppm range, or sampling line and/or instrument surface con-taminations.

Figure 4: Prominent PTR-ToF-MS ion traces (MEA, acetalde-hyde, ammonia, m/z 81.045 and one unknown peak at m/z 62.026) in the proton-transfer H3O

+ mode for 3 days duration. The plateaued profiles of all selected peaks demonstrates the capability of the PTR-ToF-MS to quantitatively monitor indus-trial emissions over an extended period. It further suggests that significant by-products caused by thermal and, more likely, oxidative MEA degradation, either require more time (days to months) to accumulate before emission to the atmosphere may be observed, or are readily produced and efficiently emitted to the atmosphere based on their physical properties. Note that the prevailing ammonia signal is out of the linear response range of the PTR-ToF-MS.

Figure 6: PTR-ToF-MS measurements performed in charge-transfer NO+ mode, for identification of unknown molecules. Sections “a, b, c” denote the periodic sampling sequence within a 120 min cycle: To Abs, From Abs, and Stripper, respectively. The tentative identification of pyrazine was sup-ported based on an observed molecular ion mass of 80.037 and its identical temporal pattern in the sampling cycles. In comparison, during the NO+ io-nisation mode measurements neither ions at m/z 60.01 ([M-H]+, H-abstrac-tion from the molecular ion) nor a molecular ion at m/z 61.02 (M+) were ob-served, and this suggests that the aforementioned species [CH4NO2]

+ did either not stem from CH3NO2 or, the ion formed underwent fragmentation under NO+ ionisation conditions. Most likely, the detection of the ion at m/z 62.026 was due to an instrument artefact.

OverviewQuantitative real-time monitoring of atomspheric monoethanolamine and byproducts emissions from industiral-scale carbon capture at Mongstad are reported using Proton Transfer Reaction-Time of Flight-Mass Spectrometry.

a. Jordan, A. et al. A high resolution and high sensitivity proton-transfer-reaction time-of-flight mass spectrometer (PTR-TOF-MS). International Journal of Mass Spectrometry 286, 122-128 (2009). b. Lindinger, W. et al. Proton-transferreaction mass spectrometry (PTR-MS): on-line monitoring of volatile organic compounds at pptv levels. Chemical Society Reviews, 27, 347-354 (1998).c. Krems, I. J. et al. The pyrazines. Chemical Reviews 40, 279-358 (1947).

Acknowledgements: We thank ACC and TCMDA for permission to present these findings.

References

Figure 5: Zoomed time profile of PTR-ToF-MS measurements in the proton-transfer H3O

+ mode. The oxidative MEA degra-dation products, ammonia, acetaldehyde and acetone, domina-ted emissions during stable operations. MEA was only emitted in significant amounts during the initiation phase, and later ef-fectively removed by double water wash stages. Emissions were generally highest from the stripper column with excepti-on of ammonia. Prominent emission of species at protonated mass 81.045 was observed (C4N2H5

+, tentatively identified as pyrazine, which can form rapidly from oxidative MEA conden-sationc, then azeotropically volatilize with water). A highly ab-undant ion at m/z 62.026 [CH4NO2]

+ alongside the protonated MEA peak requires further identification.

Figure 1. Sampling setup schematic for amine and other VOC monitoring at the Mongstad carbon capture test facility. The red flow path was primarily made of Siltek® tubing and heated to 100-130 °C to avoid condensation and minimize wall adsorption. The black flow paths were not heated but also consisted of Siltek® tubing, and the grey flow paths were made of SS or Teflon PFA (bp1). Abbreviations: bp - bypass; FC - flow con-troller; TEE - a (Siltek®) plumbing fitting in the shape of a T. The sample stream was di-luted 1:10 or 1:20 with bottled zero air and monitored in 10s intervals, later averaged to 30 s.

Figure 2. Calibration curves for toluene, trimethylbenzenes and di-chlorobenzenes, obtained from the measurements of a multi-compo-nent aromatics standard (Restek Inc., USA) with different dilution fac-tors. The quantification of those uncalibrated VOCs in absolute con-centrations as gas mixing ratios (parts per billion by volume, ppbv) was derived from data of calibrated VOCs (i.e., toluene), via applica-tion of transmission efficiency and reaction rate constant knowledge.b

Absorber and stripper stack emission measurements during a carbon capture from flue gas using MEA solution test phase at the Technology Center Mongstad

4th conference on Atmospheric Chemical Mechanisms, UC Davis (Dec. 10th-12th, 2012)

reinitialization phase reinitialization phase

no �ue gas no �ue gas